AdWords
1.6K members online now
1.6K members online now
Get started with AdWords - learn the basics to get set up for success
Guide Me
star_border
Reply

New testing framework - Day parting campaign

Follower ✭ ☆ ☆
# 1
Follower ✭ ☆ ☆

I'm trying to come up with a testing framework to reduce wasted spend. The problem with testing new bidding strategies & attributions on the same replicate of a campaign lies with cannibalizing.

 

This new campaign structure would entail having campaign A and B (B being an exact replicate of A), but having A target odd hours of the day and B target even hours. 

 

I would replicate the campaign (B), set in the odd hours under Ad Schedule and put in a -100% Bid adj accordingly. I'd do the same with campaign (A) and go -100% on all even hours.

 

The only problem is that AW only allows up to 6 time segments to be defined, but in this case I need up to 24 (one for each hour, negating either odd/even in their respective campaigns). Is there a workaround to this?

1 Expert replyverified_user

Re: New testing framework - Day parting campaign

Top Contributor
# 2
Top Contributor

Johnny;

I think a better way, with a more reliable (i.e statistically significant ) result will be to test different ad-copies using Adwords experiments;
https://support.google.com/adwords/answer/2385204?hl=en#common_goals

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: New testing framework - Day parting campaign

Follower ✭ ☆ ☆
# 3
Follower ✭ ☆ ☆
Thanks Moshe - but my test is more complex and goes beyond mere ad AB testing. For example, if you were to pit two automatic bidding strategy methods against one another or a different attribution model (last click vs time decay) this would be ideal.

Re: New testing framework - Day parting campaign

Badged Google Partner
# 4
Badged Google Partner
Johnny,

Food for Thought.

If you are trying to get every hour of the day, and only allotted 6 time sets for your request, then it's simple math to make more campaigns then Just A vs B

A vs B vs C vs D etc to define your parameters.

However I don't think you will be able to draw conclusions and will make analysis difficult adding multiple variances into the mix. I am not sure you will get enough impressions in some of the off hours and even in the 1 hour slots to really tell what is what.

Have you looked at your historical data already to determine what times are the best ROI conversions? are there times you should not be advertising? Can you limit your Range to just the times that do convert?

Depending on product and targeting area, midnight - morning hours may not be the best ROI conversions. Weekends may not be the best conversions or vice-vera. I would first look historically, then use that data to define my best testing environments.

Hope that helped, hard to give advice when not knowing all the information.

Re: New testing framework - Day parting campaign

Top Contributor
# 5
Top Contributor

Well... This is correct that Adwords experiments are a single variable experiments,

I would be cautious using a multi variable stsitsitical experiment without the assistance of  a statistician   to analyze the results.

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’